Europe 2026: heat pumps are 18% more efficient with underfloor heating than radiators

The Europe-wide efficiency gap: underfloor heating versus radiators

Underfloor-heating configurations post an average SCOP of 4.55 in the current European market snapshot, but the corpus provided here does not contain the emitter-level SCOP split needed to verify the claimed 18% gap versus radiators, so that number cannot be confirmed from the available data (market_index_snapshot).

That limitation matters because emitter choice is central to the retrofit argument. The market-wide average SCOP across heat-pump models with heating SCOP data is 4.55, with 60,989 total models in the broader EPREL snapshot and 30,452 air-to-water units alone dominating the hydronic segment (market_index_snapshot). By type, average SCOP is 4.54 for air-water, 4.77 for ground-water, and 6.15 for water-water systems, though the last group is based on just 31 models (type_efficiency).

So the broad takeaway is still directionally clear: system design around flow temperature can plausibly shift performance enough to move a project from marginal to attractive. But the exact underfloor-versus-radiator averages, model counts by emitter, and percentage gap are not present in this corpus, so they should not be stated as measured facts here. Readers wanting the wider market context can use the market index snapshot, the full heat-pump catalog, and the air-to-water catalog slice.

How big the spread is within the market, not just on average

What the data does show very clearly is that the spread inside the market is huge. Among the top 15 air-to-water models by SCOP, the best listed unit reaches 7.0, with several others clustered between 6.8 and 6.97 (top_models). At the bottom of the same air-to-water ranking, the weakest listed model is at 2.08, and multiple others sit around 2.52 to 2.70 (top_models).

That means the observed range within listed air-to-water models runs from 2.08 to 7.0, a span of 4.92 SCOP points, and the highest value is roughly 3.37 times the lowest (top_models). Even restricting the comparison to the top and bottom entries shown, the difference is much larger than the kind of gain most households get from a tariff switch or a modest controls upgrade.

The examples are concrete. At the top end, Risch Kälte- und Klimatechnik GmbH OH I 4esr TWW W/W is listed at SCOP 7.0, while Master Therm tepelná čerpadla s.r.o. AQ30I-0WW and Hoval Aktiengesellschaft 42 -Thermalia® twin (26) GW both sit at 6.97 (top_models). That upper tail is visible in the top SCOP air-to-water leaderboard, and it helps explain why blanket statements about “average heat-pump performance” are often too crude for retrofit decisions.

Which heat-pump types benefit most from underfloor systems

The corpus does not include emitter-specific SCOP by heat-pump type, so it is not possible to say which type shows the biggest underfloor-versus-radiator gap, nor to quantify that gap in points or percent from this dataset.

What can be said is that type differences alone are already meaningful. Water-water models lead the table at 6.15 average SCOP, ahead of ground-water at 4.77 and air-water at 4.54 (type_efficiency). But water-water is a niche category with only 31 models, versus 213 ground-water and 30,452 air-water entries (type_efficiency). In practice, that means most European buyers are still choosing within the much larger air-water market, where emitter choice is likely to matter most simply because that is where the volume is.

The market structure reinforces this. Air-water accounts for about half of all listed models: 30,452 out of 60,989 total, or roughly 49.9% (market_index_snapshot). Ground-water is only about 0.35%, and water-water about 0.05% of the total model pool (market_index_snapshot). For brand context inside that mass-market segment, Daikin Europe N.V. alone accounts for 14,668 models or 24.05% of all listed entries, ahead of Mitsubishi Electric Europe B.V. at 5,575 models and 9.14% (market_index_snapshot).

Where the retrofit economics work best today

A simple threshold answers one of the key running-cost questions: if gas boiler efficiency is treated as 100%, then a heat pump with SCOP 4 beats gas on fuel cost wherever the electricity-to-gas price ratio is below 4.0. On that basis, 24 countries in the price-ratio table meet the threshold, while only the United Kingdom at 4.63 and Romania at 5.11 do not; five others have no gas ratio listed (price_ratio).

The strongest running-cost positions are in countries with the lowest electricity-to-gas ratios. Sweden is the most favorable at 1.3, followed by the Netherlands at 1.49, Portugal at 1.73, France at 1.78, and Italy at 2.0 (price_ratio). In those markets, a SCOP-4 heat pump has a large arithmetic buffer versus gas before other factors such as standing charges or system losses are considered.

If the question is where wide cost advantage and high subsidies overlap, France stands out most clearly in the provided data. Its electricity price is €0.2561/kWh and gas price €0.1436/kWh, for a ratio of 1.78, while the maximum subsidy reaches €11,000 and grid intensity is just 56 gCO2/kWh (price_ratio; country_compare). Austria also combines a sub-4 ratio of 2.68 with a much larger maximum subsidy of €23,000 and a low-carbon grid at 89 gCO2/kWh (price_ratio; country_compare). Germany has a high subsidy ceiling of €21,000, but its ratio is a less favorable 3.16 and grid intensity is much higher at 366 gCO2/kWh (price_ratio; country_compare). Poland offers the highest listed maximum subsidy at €31,000, yet its ratio is 3.71 and grid intensity is the highest in the table at 661 gCO2/kWh, making the climate case weaker even where the capital support is strong (price_ratio; country_compare).

For country-level comparisons, the 32-country dashboard, France profile, Austria profile, and subsidy index are the most relevant reference points.

Why the gap matters for buyers, installers, and subsidy design

Even without the missing emitter split, the numbers show why performance differences deserve more weight in retrofit planning. A market with an average SCOP of 4.55 but an observed air-water range from 2.08 to 7.0 is not a market where “a heat pump is a heat pump” (market_index_snapshot; top_models). Small differences in design temperature, emitter compatibility, and hydraulic setup can compound into large differences in annual running cost.

For buyers and installers, that means the right comparison is not just model versus model, but system versus system. A unit chosen from the heat-pump catalog or the leaderboards hub still has to be matched to emitters and climate. That is especially important in countries where price ratios are near the SCOP-4 threshold, such as Belgium at 3.9 and Poland at 3.71, because a performance shortfall can quickly erase a nominal cost advantage (price_ratio).

For policymakers, the same logic points to subsidy design that rewards low-temperature-ready retrofits, not just appliance swaps. The corpus supports the economics side of that argument in countries such as France and Austria, where running-cost ratios are favorable, subsidies are substantial, and grid CO2 is low (price_ratio; country_compare). What it does not support, from the data supplied here, is a quantified Europe-wide 18% emitter gap. That claim needs an emitter-level EPREL extract before it can be published as measured fact.

Sources

type_efficiency — EPREL Public API · type aggregation. Snapshot: 2026-07-12.
market_index_snapshot — Househeating Pulse · Market Index v1, computed from EPREL Public API. Snapshot: 2026-07-12.
top_models — EPREL Public API via Househeating Pulse catalog. Snapshot: 2026-07-12.
country_compare — Eurostat · NASA POWER · EEA · Househeating Pulse subsidy register. Snapshot: 2026-07-12.
price_ratio — Eurostat household band DC (electricity) / D2 (gas), latest semester. Snapshot: 2026-07-12.